Retirement of Xalan-C
a. Retire the project to the Attic (My Vote)
b. Keep the project going
I appreciate the work of Roger Leigh, William Blugh, and Gary Gregory on
Xalan-C as a project with little usage. My major work with xalan-c was
the release of 1.11. I don't have the time to mentor and implement
features such as:
* Caching standard documents instead of requiring internet access to
W3 repositories to support validation in absence of W3 internet access.
* Implementing additional date-time extensions based on a time
continuum number and adding a robust namespace plugin capability.
* Ensuring that the code base supports STL deprecation of features in
new C++ language implementations. This requires implementing new
templates in the low-level code base.
* The project relies on Xerces-C for the XML readers.
We don't have the contributors and committers for these issues. The
health of the related xerces-c project is also an issue.
The various commercial projects I helped maintain have discontinued
using xalan-c in their design and maintenance. I don't know of other
projects that incorporate xalan-c as a requirement.
Linux Distributions
* Fedora (epel) continues to have xalan-c 1.12.0 for Centos-8-Stream
and RHEL-8.
* Debian-10 distributes the old xalan-c 1.11
The mail lists have seen virtually no traffic regarding the use and
issues of Xalan-C.
As for going forward, my vote is for (a).
a. Retire the project to the Attic
b. Keep the project going
Sincerely,
Steven J. Hathaway
On 10/15/2022 4:42 AM, Gary Gregory wrote:
Retirement of Xalan-C seems ok to me if only due to my lack of
involvement with it; I've only helped on the Java side IIRC. So that
would be (a) for me.
Gary
On Fri, Oct 7, 2022, 08:19 Roger Leigh <rle...@codelibre.net> wrote:
Dear all,
It’s been over three months since my original email on this
subject. There is a related discussion about this on the
Xerces-C++ mailing list just now, and it would be useful to reach
a conclusion on this for Xalan-C as well.
I've updated the git statistics I did earlier in the year, which
can be viewed or downloaded here: xerces-xalan-git-monthly.xlsx
icon xerces-xalan-git-monthly.xlsx
<https://codelibreconsulting.sharepoint.com/:x:/s/Opensourcesoftware/EabAzxgzU3pCjUSKSVvWjZgBlUGZUb91q2PVMkGk1oaIHw?e=MVBvPA>.
There are no changes—there has not been a single commit to the
source repository since 2021. There has not been any change to
the maintenance status of the project since my last email: there
are no active maintainers, no one has shown any interest in doing
any maintenance, and none of the previous maintainers who are
still present actually use Xalan any longer—so there is little
prospect of previously active maintainers returning. I myself
will be leaving the project once this question is answered
irrespective of the outcome—I no longer use Xalan-C, I have no
time to commit to it for future work and releases, I just want to
see it retired gracefully so that we don’t leave anyone with the
mistaken impression that this is a project which is active and
well supported when it is most certainly not. This is not a
library which new projects should be considering to use.
This is the commit history since 01 Oct 2012:
$ git shortlog -s --oneline --all --since "01 OCT 2012"
1 Benjamin Beasley
1 Bill Blough
1 Biswapriyo Nath
1 Kvarec Lezki
182 Roger Leigh
29 Steven J. Hathaway
I would like for the PMC to vote on the future of the project. Do we
1. Retire the project to the Attic
2. Keep the project going
I’m not sure if I’m formally a PMC member or not, but
realistically I’m the only one who has done any work on the
project for the past 8 years. So if I can vote on this I’ll vote
for (a).
Kind regards,
Roger
*From:*rle...@codelibre.net <rle...@codelibre.net>
*Sent:* 22 June 2022 23:21
*To:* d...@xalan.apache.org; c-users@xalan.apache.org
*Subject:* Future of xalan-c
Dear all,
I wanted to write this email to sound out where the project is,
where it is going, and whether or not it has a future. If it does
not have a future, is it time to wrap up the project and move it
to the Attic?
To start with, a bit of context. This is a summary of the
project’s commit activity over the previous 22 years:
Back in July 2020, just a little under two years ago, I released
Xalan-C 1.12. This was the first release since Xalan-C 1.11 in
October 2012, and it incorporated a number of patches which had
been accumulated over the course of years by several downstream
distributors.
https://apache.github.io/xalan-c/releases.html#major-changes shows
the major changes in this release. On the above graph, this
release is comprised of the commits from 2019 to 2020. I was the
/sole/ committer for this release.
The previous 1.11 release was made in October 2012 with Steven J.
Hathaway being the principal contributor.
The previous 1.10 release was made in October 2005 with David N
Berton and Dmitry Hayes being the principal contributors.
The previous 1.9 release was made in December 2004 with June Ng,
Matthew Hoyt, David N Berton and Dmitry Hayes being the principal
contributors.
The previous 1.8 release was made in April 2004 with Matthew Hoyt,
David N Berton and Dmitry Hayes being the principal contributors.
The main points I’d like to make here are the following:
* Active development of Xalan-C effectively finished with the
/1.10/ release in 2005. The vast majority of work since then
has been little more than essential bugfixing and portability
work to support new platforms and toolchains.
* 1.11 was a bugfix release. It was primarily comprised of
essential bugfixes, and fixes for building with different
toolchains on different platforms and some documentation
work. There was one code improvement of note: “Add number and
nodeset types as top-level stylesheet parameters”
* 1.12 was a bugfix release. It was primarily comprised of
essential bugfixes, and fixes for building on different
platforms, with the CMake support generalising that to build
on current platforms, plus the documentation switch to
Markdown. There were zero new features or improvements outside
essential bugfixing.
* There is essentially ~zero developer mailing list activity
* There is essentially ~zero user mailing list activity
* Community involvement on GitHub is present but at very low and
sporadic levels. We have three PRs from contributors other
than myself
(https://github.com/apache/xalan-c/pulls?q=is%3Apr+is%3Aclosed).
One was a triviality, two were portability fixes just altering
platform-specific ifdefs. There is one open PR
(https://github.com/apache/xalan-c/pulls?q=is%3Aopen+is%3Apr).
This looks simple but I’m not sure of the impact in case of
unexpected subtleties.
I became involved in the project for pragmatic reasons—I worked on
a project using XSLT and picked up Xalan-C as a dependency. I
wrote and contributed the CMake support and worked on the 1.12
release for that reason. But I don’t know the underlying
codebase, and I can’t do any real feature development or deep
bugfixing. I don’t have the expertise with XSLT, or the time to
do this. And since I no longer work on any projects using
Xalan-C, I’m no longer realistically able to do any further
maintenance work either. If I hadn’t done the most recent work
and made the 1.12 release, it’s most likely that the incorporation
of community patchsets and making a point release would not have
happened. No one aside from me has worked on Xalan-C since Steven
J Hathaway’s last work in 2012.
I don’t personally think there is sufficient community involvement
or developer involvement to realistically support Xalan-C as an
active project in any sense. There is no one working on it. And
while I’m sure there are some users, there’s next to no active
engagement of users as a community.
I’ve made a good effort to keep the project going for the near- to
medium-term. The CMake build made it possible to build on all
contemporary platforms. The documentation switch to Markdown made
it possible to build without obsolete and unavailable Java
libraries. The bugfixes we included in 1.12 fixed a number of
critical issues. So 1.12 should serve as a usable release for the
foreseeable future even in the absence of further development.
However, I don’t see a future for anything beyond 1.12 unless
there is a dramatic change. XSLT usage is declining, and Xalan-C
doesn’t support XSLT 2.0 and beyond. Rather than letting the
current situation linger on indefinitely, I wanted to suggest we
take stock of where we are, and if there is consensus to do so, I
think it would be advisable to draw a line at this point and end
the project gracefully.
Kind regards,
Roger
